Yuxiang Qin1, Alan R. Duffy2, Han-Seek Kim1, Simon J. Mutch1, Gregory B. Poole1,2,
Paul M. Geil1, Andrei Mesinger3 and J. Stuart B. Wyithe1
1School of Physics, University of Melbourne, Parkville, VIC 3010, Australia
2Centre for Astrophysics and Supercomputing, Swinburne University of Technology, PO Box 218, Hawthorn, VIC 3122, Australia
3Scuola Normale Superiore, Piazza dei Cavalieri 7, I-56126 Pisa, Italy
abstract
In making direct comparisons between hydrodynamic and semi-analytic galaxy formation models, which start from the same cosmological initial conditions, we studied gas accretion, cooling and star formation of dwarf galaxies at high redshift (z>5) by isolating feedback in the companion paper. We proposed several modifications to semi-analytic models for more accurate predictions of high-redshift dwarf galaxies. In this work, we extend the comparison with reionization and supernova feedback included, as well as two semi-analytic star formation prescriptions based on the consumption of the entire cold gas disc and the molecular hydrogen only. We focus on galaxies with 1e9M⊙ < Mvir < 1e11M⊙. We find that with the modifications previously proposed, including to suppress the halo mass and baryon fraction as well as modulate gas inflow and star formation efficiencies, the current SAM is able to reproduce the cosmic evolution of galaxy properties predicted by the hydrodynamic simulation. These include the galaxy stellar mass function, total baryonic mass, cold gas mass and star formation rate at z=5-11. However, in order to more accurately model dwarf galaxies with 1e9M⊙ < Mvir <1e10M⊙ using the cold gas-based star formation law, we find that the threshold of the cold gas mass for star formation needs to be reduced. Otherwise, a significant cold gas will be stored in dwarf galaxies that have been quenched.
Key words: galaxies: formation — galaxies: dwarf — galaxies: high-redshift — method: numerical